Influence of Local Communities on Organizational Behavior

Transcription

1 Acting Globally but Thinking Locally? The Influence of Local Communities on Organizations Christopher Marquis Julie Battilana Copyright 2007 by Christopher Marquis and Julie Battilana Working papers are in draft form. This working paper is distributed for purposes of comment and discussion only. It may not be reproduced without permission of the copyright holder. Copies of working papers are available from the author.

2 Acting Globally but Thinking Locally? The Influence of Local Communities on Organizations Christopher Marquis Harvard Business School Boston, MA Julie Battilana Harvard Business School Boston, MA Prepared for Research in Organizational Behavior November 27, 2007 We thank Joe Galaskiewicz, Royston Greenwood, Arndt Sorge, and Jean-Claude Thoenig for their insightful comments and suggestions on a prior draft of this paper.

3 Acting Globally but Thinking Locally? The Influence of Local Communities on Organizations ABSTRACT We develop an institutional theory of how local communities continue to matter for organizations, and why community factors are particularly important in a global age. Since globalization has taken center stage in both practitioner and academic circles, research has shifted away from understanding effects of local factors. In this paper, our aim is to redirect theoretical and empirical attention back to understanding the determinants and importance of local influences. We review classical and contemporary research from organizational theory, sociology and economics that have focused on geographic influences on organizations. We adapt Scott s (2001) influential three pillars model, including regulative, social-normative and cultural-cognitive features to conceptualize an overarching model of how communities influence organizations. We suggest that because organizations are simultaneously embedded in communities and organizational fields, by accounting for both of these different levels, researchers will better understand isomorphism and change dynamics. Our approach thus runs counter the idea that globalization is a homogeneity-producing process, and the view that society is moving from particularism to universalism. With globalization, not only has the local remained important, but in many ways local particularities have become more visible and salient, and so understanding these dynamics will be helpful for researchers addressing institutional isomorphism and change. 2

4 Acting Globally but Thinking Locally? The Influence of Local Communities on Organizations It is a paradox of recent times that in a globalizing and boundaryless economy, factors associated with local communities are of central importance to understanding organizations and their actions (Bagnasco and Le Galès, 2000; Sorge, 2005; Marquis, Glynn and Davis, 2007). Recent studies have shown that embeddedness in communities has an enduring influence on organizational behavior and there are a number of mechanisms that mediate this relationship. For example, geographic proximity and local networks influence organizations non-profit giving (Galaskiewicz, 1997), board of director structure (Kono et. al, 1998; Marquis, 2003) and corporate governance practices (Davis and Greve, 1997). There is also evidence that different localities exhibit shared frames of references which influence outcomes as diverse as corporate social responsibility behaviors (Marquis, Glynn and Davis, 2007), corporate strategies (Lounsbury, 2007), governance processes (Abzug and Simonoff, 2004) and organizational foundings (Marquis and Lounsbury, 2007). Variation in local laws and tax rates also contribute to differences in organizational behavior across communities (Guthrie and McQuarrie, 2005; Guthrie et al, forthcoming; Marx, Strumsky and Fleming, 2006). Further, a growing stream of research focuses on how local competitive and market-based processes influence organizations (Audia, Freeman and Reynolds, 2006; Stuart and Sorensen, 2003; see Freeman and Audia, 2006 for a review). This diverse work suggests that even in spite of recent globalizing trends, there has been a revival of research accounting for the effect of geographic communities on organizational 3

5 behaviors. But there has not yet been a theoretical synthesis that delineates scope and boundary conditions as well as the underlying processes that drive these relationships. Our approach and review of how communities influence organizations will focus on how the tools and mechanisms of institutional theorizing (Scott, 2001; Davis and Marquis, 2005) can enhance our understanding of the influence of local forces in a global age. The primary underlying premise of institutional theory is that action and choice cannot be understood outside of the cultural and historical frameworks in which organizations are embedded, yet paradoxically, the theory has thus far mostly neglected the important influences that are associated with organizations local cultures, legal systems and social contexts. While early institutional works, such as Selznick s (1949) study of the Tennessee Valley Authority and Zald s (1970) study of Chicago YMCA were heavily influenced by local sources of power, as Powell and DiMaggio (1991) describe, institutional theory has more recently discarded the focus on local environments to more frequently focus on geography-independent organizational sectors, or fields. The organizational field (DiMaggio and Powell, 1983) has proven to be a powerful level of analysis which has shifted research attention and obscured the influence of geographic embeddedness. Our goal is to build on the social constructionist and cognitive tradition of institutional theory, but reorient the theory to focus more on community influences. Our review of this literature harks back to some of the earliest work in modern organizational theory that focused on documenting the importance of community in understanding organizational behavior. Early investigators such as Warren (1967), who, following upon Emery and Trist (1965), coined the concept of inter-organizational field, stressed the importance of community for understanding institutional influences 4

6 because it is within communities that diverse types of organizations come into contact. Other early research documented how social linkages develop among diverse organizations located within same community (Litwack and Hylton, 1962; Turk, 1977; Lincoln, 1979). Ironically, much of this early work sought to demonstrate that the American community was undergoing great change, which involved the increasing orientation of local communities units toward extra-community relations. In explaining why community studies had fallen out of favor after the 1980s, Scott (2001) describes that modern transportation and communication systems developed such that geographical boundaries became meaningless. More recently, globalization has come to occupy the center of the stage, further contributing to the neglect of the community level of analysis (Sorge, 2005). But the recent flurry of scholarship and research on community bases of organizational behavior suggests that Scott s (2001) assessment is not accurate. The most systematic of this research has been in the ecological tradition, building of Hawley s (1950) community ecology model, as adapted by Hannan and Freeman (1977, and see Freeman and Audia, 2006 for a review). These studies that highlight competitive and ecological factors of geographic communities are important and focus on how proximity defines market boundaries and as a result still drives some organizational decisions. As a counterpoint, our focus is on how communities are not just contexts for competition, but provide different institutional environments which influence organizations. We argue in this paper, that communities influence organizational behavior not only as local markets and resource environments, but also through a number of institutional pressures. While a stream of research in economic geography (e.g., Storper, 5

7 1997; Scott and Storper, 2003) has examined the influence of some institutional factors on the economic development of communities, it is striking to note that recent developments in institutional theory have tended to overlook the influence on organizational behavior of institutional pressures stemming from the community. As an analytic device to unpack the different institutional influences of communities, we follow Marquis, Glynn and Davis s (2007) research on the geography of corporate social responsibility and draw on Scott s (2001) three pillars model, including regulative, social-normative and cultural-cognitive features of communities. In doing so, we hope to reorient institutional theory from the current focus on organizations embeddedness in organizational fields to organizations simultaneous embeddedness in both geographical communities and organizational fields. Building on Warren s (1967) insight, we feel that accounting for community-level processes will draw the social and cultural underpinnings of organizational behavior into fuller relief by showing how even organizations with conflicting economic purposes are influenced by embeddeddess in similar geographic environments. By accounting for these different levels of analysis, we also believe that researchers will better understand isomorphism and change dynamics both within and across geographical communities and organizational fields. While communities may be a natural venue to understand isomorphism processes (Marquis, Glynn and Davis, 2007), there is evidence as well that as different local communities come closer together as a result of increased globalization, they may also demarcate the boundaries between them even more clearly (Scott and Storper, 2003; Sorge, 2005; Marquis and Lounsbury, 2007). Such an approach to the process of globalization runs counter the idea that one can view 6

8 the evolution of society as moving from particularism to universalism through homogeneity-producing trends (Robertson and Khondker, 1998; Sorge, 2005). In closing, the goal of our chapter will be to both review the emerging set of work that has considered communities effects on organizations and further to develop a theoretical synthesis that delineates scope and boundary conditions as well as the underlying mechanisms that drive these relationships in the context of a globalizing economy. After defining what we mean by community, we review some of the geographically-oriented work in organizational ecology and economics that focuses mainly on effects of proximity and studies communities as competitive environments. To complement this work and better account for institutional pressures stemming from the community, we then apply Scott s (2001) influential typology of institutional features to the community level of analysis. We conclude with some final thoughts about how understanding communities may be even more important in light of globalization and future research directions and extensions suggested by taking this approach. Focusing on communities as institutional environments provides fresh theoretical insights to organizational theory in addition to providing a more unified perspective on this diverse set of emerging community-oriented research. LOCAL COMMUNITIES: A NEGLECTED LEVEL OF ANALYSIS Many sociological and anthropological definitions of the concept of community exist, and most emphasize some combination of relative small-scale, boundedness, and strong ties among the members of the community (Oxford dictionary of social sciences, 2001). These qualities distinguish community from larger and more impersonal forms of relationship such as society, as emphasized in Tönnies's (1887) seminal distinction 7

9 between Gemeinschaft (community) and Gesellschaft (society) whereby community is about collective relationships between people focused on interpersonal and particularistic connections, and society is more universal, transparent and anonymous. In his influential work on economic geography, Storper (2005:34) draws on this distinction and defines community as referring to a wide variety of ways of grouping together with others with whom we share some part of our identity, expectations, and interests. Although the distinction between community and broader society is helpful, and it highlights many of the important components that comprise community, these definitions do not precisely delineate the boundaries of a local community. Such boundaries are hard to delineate in an abstract definition and need not necessarily coincide with any political or administrative boundaries. Warren (1967: 400) explains, the term community level does not imply a discretely identifiable level, except for purposes of analysis. Using the term community field, he thereby intertwined the concepts of field and community and showed how even organizations with conflicting economic purposes are influenced by embeddedness in similar geographic environments. In conceptualizing communities, we think it is important to highlight our focus on bounded geographic entities in order to effectively distinguish our meaning from definitions of geography-independent organizational fields that have emerged as the unit of analysis used in institutional studies to account for the wider institutional context in which actors are embedded (Davis and Marquis, 2005). Further, a focus on geographic boundaries serves as a counterpoint to recent institutional research that aims to understand the importance of transnational phenomena (e.g., Djelic and Quack, 2003; 8

10 Djelic and Sahlin-Andersson, 2006) in a world where organizations are more open to non-local events and ideas (Scott, 2005: 474). There is significant precedent for defining community as a metropolitan region. This is the approach taken by early American sociology and political science, in studies of Muncie, Indiana (Lynd and Lynd, 1929), Newburyport, Massachussetts (Warner and Lunt, 1941), Philadelphia, Pennsylvania (Baltzell, 1958), Atlanta, Georgia (Hunter, 1953) and New Haven, Connecticut (Dahl, 1961; Polsby, 1963). Recent research (e.g. Marquis, 2003; Stuart and Sorenson, 2003) has followed a similar approach. An early European tradition of research also addressed the importance of cities as essential social structures, which were local societies where groups and interests gathered and were represented (Weber, 1921). Weber proposed to analyze the city through its economy, its culture and its politics, which were interconnected. Anchoring itself in the Weberian tradition, a more recent stream of research has revived interest in European cities as local units of analysis, which paradoxically remain significant tiers of social and political organization, in the era of globalization (Bagnasco and Le Galès, 2000: 6). But we argue that the phenomena associated with communities need not be bound within a city-limit. Aldrich (1999: 300) for example, suggested that the geographic scope of a community is an empirical question. He stressed that what qualifies as a community is to some extent determined from the bottom-up based on relations between organizations. Thus, other geographic entities, for example clusters of cities or regions may also qualify as communities (Greenwood, Diaz, Li and Lorente, 2007). Economic geography has contributed to the resurgence of regions as units of analysis in social sciences over the last two decades by examining their role in economic development and 9

11 considering some of them as sites of the most advanced forms of economic development and innovation (Scott and Storper, 2003). Silicon Valley (Saxenian, 1994) is a good example of a region playing an active role in the development and the improvement of industrial products and processes. Recent research has focused on how industrial regions become identified by external audiences which results in a reinforcing process (Romanelli and Khessina, 2005). Industrial districts (the term was coined by Alfred Marshall (1920)), which are geographically localized industrial systems based on an extended division of labor between small and medium-sized firms that, although they often directly compete with each other, also cooperate with each other in a number of different ways (Piore and Sabel, 1984; Brusco, 1995), also qualify as local communities. They are industrial systems built on regional networks. More broadly, any local productive system defined as a system composed of three principal elements: the active businesses, the territory in which they are located and the people living in that territory, with their values and their history (Brusco, 1995: 63) qualifies a geographic community. We regard the community level of analysis as a local level of analysis corresponding to a the populations, organizations and markets located in a geographic territory and sharing, as a result of their common location, elements of local culture, norms and identity. We recognize that the delineation of the boundaries of such territory is not straightforward. The boundaries of local communities are not given; they are always partially constructed by researchers in the same way as organizational fields boundaries are. We, however, argue that the community level of analysis has to be revived in institutional theory as it is the only way to account for the fact that organizations are locally embedded. 10

12 In this paper, we focus on the underlying regulative, social and cultural mechanisms that influence organizations within local communities. By our definition, these could include cities, clusters of cities, regions, industrial districts or any type of local productive systems. After reviewing studies that focus on how geographic proximity and local markets influence organization behaviour, as a counterpoint, we highlight geographic communities influence from an institutional standpoint. We thereby suggest that communities are an essential level of analysis in understanding the interactions between organizations and their environment. COMMUNITIES AS LOCAL MARKETS The most developed streams of organizational research that has explicitly studied local effects on organizations are those that examine local competitive environments that provide differential levels of various types of resources for organizations. In this stream of research, investigators identify organizational communities based on geographic proximity and study the effects of local markets on organizations economic performance. We focus here on two influential streams of research. First, we review the contribution of the ecological tradition, which builds on the work of Hannan and Freeman s (1977), whose original population ecology model theorized the importance of communities for understanding the dynamics of organizational populations (see Freeman and Audia, (2006) for a review of this work). Secondly, we review the work on economic geography (Marshall, 1920), which examines how geographic collocation of industries provide positive externalities, such as spillovers and labor training. Some researchers, for example Sorensen and Audia (e.g. Sorensen and Audia, 2000; Audia, 11

13 Freeman and Reynolds, 2006) are increasingly integrating insights from both of these approaches. The ecological research stream mostly focuses on how organizational density in a population influences the vital rates of other organizations, i.e., organizations die and are founded as a function of the existing stock of similar organizations. The key theoretical advancements of this approach, density dependence and resource partitioning, both have been shown to be geographically contingent and in some cases more dependent on local market processes than those of the entire field (Freeman and Audia, 2006). For example, Carroll and Wade (1991), in a study of how local and national densities in the brewing industry influence population dynamics find that local competition matters more than national competition. The most developed theoretical consideration of local environments is Greve s (2000, 2002) spatial density dependence model, which posits that localized competition is more central than field-level characteristics to organizational decision making. Markets have bounds, and some types of organizations, in Greve s case, Tokyo banks, live within those bounds. Further, Baum in a number of studies (Baum and Singh, 1994; Baum and Mezias, 1992) has also shown how localized competition and crowding within local communities led to greater levels of failure rates for community day care centers in Toronto and Manhattan hotels. This line of work has shown that collocation and proximity are important to defining organizational ecologies and sites of market competition. The ecological model of resource partitioning theorizes a relationship between the consolidation of markets and founding of new firms. As a market consolidates into a fewer number of generalist firms, specialist firms arise to capitalize on market niches 12

14 abandoned by the larger competitors. Traditionally, the explanation has focused on proximate location: an organization s location in the resource space accounts almost entirely for the partitioning of industries (Carroll and Swaminathan, 2000). McPherson (1983) proposes a similar model whereby overlap of member characteristics defines niches and competition between voluntary organizations. In most cases, the site of resource competition in these studies is cities. For example, Carroll s (1985) original statement of resource partitioning studied newspapers in seven US cities because they were the autonomous units of competition for newspapers. Marquis and Lounsbury (2007) also find support for resource partitioning as a community process in their study of how local bank acquisitions lead to new bank foundings. Further supporting the effects of local mergers on organizational founding, Stuart and Sorenson (2003) found that organizational liquidity events such as IPOs and acquisitions within focal or adjacent communities leads to the foundings of technology firms. These effects of density dependence and partitioning of local markets, which focus on some of the negative effects of crowded or consolidated markets, are to some extent in conflict with other research in the economic geography area that study how industries agglomerate and how close geographic proximity with competitors can be beneficial (Marshall, 1920; Krugman, 1991). Ecology suggests there are negative effects of crowding as resources become scarcer. But work in economic geography suggests an opposite relationship. Focusing more on the local accumulation of knowledge, and trained labor, which leads to information spillovers, it shows that there is a substantial benefit to all local firms from the agglomeration of industries. For example, Silicon Valley became the center of the technology industry as a result of tight networks of local 13

15 firms (Saxenian, 1994), Detroit became the capital of the auto industry by having fertile early training grounds such as the Olds company that spawned many spin-offs (see Klepper, 2002 for summary of his other research) and Akron was fertile soil for tire research leading to benefits for firms located and founded there (Sull, 2001; Buenstorf and Klepper, 2005). These studies begin accounting for some more of the social dynamics that are boundary conditions for a strict ecological approach. For example entrepreneurs are not necessarily randomly distributed, and individuals typically start businesses in close proximity to their current places of residence (Marquis and Lounsbury, 2007; Sorenson and Audia, 2000; Thornton and Flynn, 2003), sometimes founding a company to compete with a former employer (Burton, Sorenson and Beckman, 2002; Romanelli and Schoonhoven, 2001). Networks also have been found to lead to increased founding rates in communities (Audia, Freeman and Reynolds, 2006). In a particularly influential study, Saxenian (1994) described how the characteristics of two technology communities, Boston and Silicon Valley influenced innovation and production within these regions. Local factors such as universities, business associations, clubs and professional organizations sustained the region s culture of embeddedness. While competition and market processes are important mechanisms to understand the effects of proximity and collocation, studies such as Saxenian s suggest that local systems and organization behavior within them may in fact be better characterized by institutional explanations. 14

16 COMMUNITIES AS INSTITUTIONAL ENVIRONMENTS We believe that while the above studies have been influential, they are in many ways only a starting point in understanding how embeddedness (Dacin, Ventresca, and Beal, 1999) in communities influences firm behaviors. As noted, their conception of community is mainly about proximity and local markets. There are still significant open questions with regard to how embeddedness in a community influences organizational behavior and characteristics beyond organization performance, foundings and death. A further distinction between our approach and the community ecology approach is that while they are examining dynamics of populations, we are interested in understanding organizational behavior at a more micro-level, including how the specific behaviors and strategies of organizations are influenced by their communities. Following Marquis, Glynn and Davis s (2007) research on corporate social responsibility in drawing on Scott s (2001) influential typology of institutional processes, we argue that communities influence organizational behavior through three primary mechanisms. We start with coercive pressures, which stem from the regulative structures of the community (i.e., the formal rules and incentives constructed by empowered agents of the collective good) that may force organizations to adopt specific managerial practices or organizational forms. This is consistent with the diverse research across levels of analysis that has suggested the political boundaries are important for understanding organizations (e.g. Dobbin, 1994; Wade, Swaminathan and Saxon, 1998, Guthrie, 2003). But we also argue that organizational practices or forms may be influenced by social-normative processes, in which organizations conform to other actors expectations to obtain their approval. Finally, cultural-cognitive processes may 15

17 influence organizational behavior within communities by imposing abstract rules associated with the structure of cognitive distinctions and taken-for-granted understandings. We see the cultural-cognitive influences as distinct from the socialnormative in that the cultural-cognitive gives actors a deeply shared frame of reference that does not need action to maintain or recreate (Douglas, 1986; Berger and Luckmann, 1966). In contrast, the social-normative is more about how consensus about what is appropriate arises out of the action of collectives and from one s peers. Thus, while the cultural-cognitive is about how things are done around here, the normative has more of an evaluative tone what is appropriate to do around here. We use these three categories of processes below to review the existing community-oriented literature and unpack the various mechanisms that connect organizational behavior to community-level processes. Regulative Influence of Communities Communities exert a regulative influence on organizations. In Scott s formulation (2001: 35),...regulative processes involve the capacity to establish rules, inspect or review others conformity to them, and, as necessary, manipulate sanctions rewards or punishments in an attempt to influence future behavior. In translating this to the community level, we focus on how local politics and government mandates can influence organizational behavior within communities. First, we highlight that regulative pressures vary across communities by providing examples of such variation. Second, we analyze how different kinds of local public policies may have a determining influence on organizational behavior within communities. Finally, we show how local public 16

18 authorities may also mobilize other local actors to indirectly influence organizational behavior within a community. Variation in regulative pressures. Following Weber s definition, states are usually defined as having the monopoly of the legitimate use of violence within the national territory, thereby being the source of most regulative pressures. States, however, vary in their degree of centralization. The more decentralized states are, the more room for agency local public authorities have. Hence, the influence of local public policies on organizational behavior will vary from one country to another depending on the degree of political decentralization (Zeitlin, 1995). For example, examining the conditions that facilitate the development of industrial districts, scholars (Trigilia, 1992; Benton, 1992; Ganne, 1995) have found that in countries where political decentralization has enhanced the autonomy and powers of regional governments, such as in Italy in the 1970s and in Spain in the 1980s, local public policies may contribute to the development of industrial districts. In contrast, where the financial and political independence of local authorities are sapped by central government controls, as in contemporary Britain, industrial districts cannot flourish (Zeitlin, 1995). Types of local public policies. Local public authorities may influence organizational behavior through a variety of regulative pressures, including legal regulations as well as the creation of incentives and administrative bodies in charge of supporting different types of organizations. Local public policies based on incentives that are likely to influence organizational behavior include, among others, subsidies to industry, tax breaks, infrastructure provision, and labor training program. For example, when local governments control tax laws, they may use them to influence organizations 17

19 practices in certain sectors. Guthrie, Arum, Roksa and Damaske (Forthcoming) examine how local state tax laws shape corporate giving to local schools in the United States. They find that there is an association between higher corporate tax rates and corporate giving to local institutions, such as local schools. States with higher corporate tax rates not only create incentives and opportunities for tax write-offs thereby encouraging corporate giving (Bakija and Slemrod, 1996; Bakija and Steuerle, 1994; Bakija et al., 2003), but they also signal to corporations the importance of supporting local social services and the provision of local public goods (Guthrie et al. Forthcoming). The literature on agglomeration economies also illustrates the impact that local public policy may have on organizational behavior (Piore and Sabel, 1984; Bagnasco and Sabel, 1995). Local public authorities often play a facilitating role in the development of industrial districts. For example, in regions where industrial districts appeared in Italy, local governments have often played a facilitating role by providing a certain quantity of collective goods which have reduced costs for employers and contributed to protecting workers, thereby encouraging local compromise (Trigilia, 1995). These goods include social services for workers (transport, public housing, schools, daycare centers) and, for local firms, the provision of industrial estates, infrastructures, professional training, and support to consortia for marketing or export facilities. A particularly important legal difference that has been shown to influence innovation in technology centers is the variation in the enforceability of noncompete clauses across locales (Stuart and Sorensen, 2003; Marx, Strumski and Fleming, 2007). In addition, the local government also acted as an agent of community sentiment in limiting the violations of health and safety standards in the new shops (Piore and Sable, 1984: 229). 18

20 Similarly, Semlinger (1995) describes the role of the local government in the economic development of the German region of Baden-Württemberg since the early 1970s, when the promotion of inter-firm cooperation among small firms has been made a policy issue in the community. Local authorities have provided in-kind support for cooperative activities as well as subsidies for brokerage services, for traveling expenses in connection with network meetings, and for up to 50 percent of the costs incurred by the development of joint projects. This local public policy has stimulated hundreds of inter-firm networks, thereby contributing to the development of an industrial district. Similarly, Guthrie and McQuarrie (2005) recount how there is considerable variation across US states and cities on corporate support of low income housing. While the US federal tax credit designed to stimulate such housing is the same across the entire country, innovative cities and states have in effect developed local institutional fields, and dense networks centered on helping corporations, banks, municipalities, and neighborhood organizations utilize the credit, thus creating considerable differences across locales. Interaction with other local actors. Finally, local public authorities may also mobilize other local actors to help them shape economic and organizational behavior. Differences in the type of local actors that public authorities mobilize as well as in the action of these local actors may explain variations in the regulative environment across communities. For example, universities have been identified as playing a key role in the shaping of the institutional environment (Amin and Thrift, 1992; 1994; Phelps and Tewdwr-Jones, 1998). The university campus is like the corner café where Italian artisans solve one another s problems and share or steal one another s ideas (Piore 19